INDEX

Explanations

the word "is" within phrases

oai_token-act-pair · gpt-3.5-turbo

statements that assert a truth or reality

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.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

itches

-0.71

 receipt

-0.68

 wills

-0.68

hoe

-0.64

 encounters

-0.63

untarily

-0.62

 stump

-0.62

 throats

-0.61

 customs

-0.60

ilings

-0.60

POSITIVE LOGITS

nt

0.92

olation

0.89

olate

0.86

cussion

0.83

not

0.80

 ALWAYS

0.80

ometric

0.79

rael

0.78

 suppose

0.77

NOT

0.76

Activations Density 0.332%

the word "is" within phrases

statements that assert a truth or reality

No Comments

No Known Activations