INDEX

Explanations

phrases that express deep personal reflections

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

verbs in the past tense

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 12-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_post

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_post

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

 Flavoring

-0.71

 Jarrett

-0.70

 Frag

-0.66

 Spur

-0.64

 Glob

-0.63

 reperto

-0.62

Hut

-0.58

 Rounds

-0.58

 Kills

-0.58

 folds

-0.58

POSITIVE LOGITS

ividual

0.88

 prefer

0.81

rather

0.79

resist

0.79

love

0.78

 LIKE

0.77

etheless

0.77

swer

0.77

like

0.77

apest

0.76

Activations Density 0.032%

phrases that express deep personal reflections

verbs in the past tense

No Comments

No Known Activations