INDEX

Explanations

proper nouns or terms related to rankings

oai_token-act-pair · gpt-3.5-turbo

references to rankings or comparisons of entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.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

void

-0.73

eways

-0.65

rums

-0.65

drm

-0.65

uv

-0.62

ivan

-0.62

umb

-0.61

aghan

-0.60

vous

-0.59

Shift

-0.58

POSITIVE LOGITS

 ranking

3.84

 Ranking

3.18

 rankings

2.23

ranking

2.17

 rank

2.11

 ranked

2.06

ranked

1.82

Rank

1.73

 Rankings

1.70

rank

1.62

Activations Density 0.012%

proper nouns or terms related to rankings

references to rankings or comparisons of entities

No Comments

No Known Activations