INDEX

Explanations

occurrences of the word "no"

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

the word "no" followed by a word or phrase indicating a negative situation or impossibility

oai_token-act-pair · gpt-3.5-turbo

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

pell

-0.58

 prone

-0.57

 Hera

-0.57

ciation

-0.57

rn

-0.57

combe

-0.56

RAFT

-0.56

 pron

-0.56

lycer

-0.56

 Metatron

-0.55

POSITIVE LOGITS

xious

1.03

 longer

0.97

 matter

0.97

etheless

0.95

 doubt

0.88

ct

0.85

 sooner

0.84

 conceivable

0.83

xus

0.77

 sane

0.77

Activations Density 0.099%

occurrences of the word "no"

the word "no" followed by a word or phrase indicating a negative situation or impossibility

No Comments

No Known Activations