INDEX

Explanations

terms related to various types of items or possessions

oai_token-act-pair · gpt-3.5-turbo

references to miscellaneous categories or items that do not fall under specific classifications

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

 Tanaka

-0.73

ruary

-0.72

pload

-0.70

 Lithuania

-0.67

 Boris

-0.66

 Manson

-0.65

 Cycling

-0.65

 Zuckerberg

-0.64

 Bundy

-0.64

 Teresa

-0.64

POSITIVE LOGITS

worldly

1.52

 assorted

1.18

 forms

1.10

 aspects

1.06

 factors

1.04

 considerations

1.03

 kinds

0.99

 goodies

0.99

 types

0.98

 pertinent

0.90

Activations Density 0.088%

terms related to various types of items or possessions

references to miscellaneous categories or items that do not fall under specific classifications

No Comments

No Known Activations