INDEX

Explanations

multi-word phrases featuring specific prefixes like "multi-", "anti-", "nano-", "auto-", "mono-", "anti-" and "non-" that are commonly related to technology, medicine, and other specialized fields

oai_token-act-pair · gpt-3.5-turbo

technical terms related to complex systems and design

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 6-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.6.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.6.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Darling

-0.74

 Duchess

-0.74

 Unle

-0.71

ãģ®éŃĶ

-0.70

 Marriage

-0.70

âĸ¬âĸ¬

-0.69

 Grizzlies

-0.68

REDACTED

-0.66

++++++++++++++++

-0.66

HRC

-0.66

POSITIVE LOGITS

agic

1.04

rop

1.04

olitics

1.01

heres

1.00

resent

0.98

ublic

0.95

riend

0.94

sych

0.93

atform

0.93

unct

0.93

Activations Density 0.052%

multi-word phrases featuring specific prefixes like "multi-", "anti-", "nano-", "auto-", "mono-", "anti-" and "non-" that are commonly related to technology, medicine, and other specialized fields

technical terms related to complex systems and design

No Comments

No Known Activations