INDEX

Explanations

words related to observation and speculation about actions or events

oai_token-act-pair · gpt-3.5-turbo

pronouns and related expressions of perception or belief

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

 Effective

-0.66

é£

-0.66

 Restoration

-0.62

 Affordable

-0.61

-0.60

 Mercenary

-0.60

 Chong

-0.60

 monog

-0.59

Vi

-0.58

Fug

-0.58

POSITIVE LOGITS

 wonder

0.94

 wondered

0.88

 searched

0.87

 watched

0.84

 hear

0.84

 realise

0.83

 wish

0.82

 pity

0.81

 overlook

0.80

 admire

0.80

Activations Density 0.553%

words related to observation and speculation about actions or events

pronouns and related expressions of perception or belief

No Comments

No Known Activations