INDEX

Explanations

words related to composition or configuration

oai_token-act-pair · gpt-3.5-turbo

occurrences of 'pos' and variations relating to position or placement

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 casc

-0.70

ãģ®éŃĶ

-0.67

ARD

-0.66

loo

-0.66

eely

-0.65

ARDS

-0.65

 diminishing

-0.64

 braking

-0.63

ħĭ

-0.62

 TOUR

-0.61

POSITIVE LOGITS

pos

1.41

itory

1.22

itions

1.11

itional

1.03

idon

0.99

itives

0.98

itionally

0.97

position

0.95

itiveness

0.92

itor

0.90

Activations Density 0.007%

words related to composition or configuration

occurrences of 'pos' and variations relating to position or placement

No Comments

No Known Activations