INDEX

Explanations

instances of words related to taking specific actions or measures

oai_token-act-pair · gpt-3.5-turbo

references to actions or measures taken

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

inately

-0.84

ILLE

-0.69

olls

-0.69

orf

-0.68

bid

-0.67

gdala

-0.66

raid

-0.65

stown

-0.65

ews

-0.64

inite

-0.64

POSITIVE LOGITS

iblings

1.02

 steps

0.97

hooting

0.94

 Steps

0.87

hops

0.86

hent

0.81

steps

0.78

 forward

0.78

isters

0.76

 toward

0.76

Activations Density 0.036%

instances of words related to taking specific actions or measures

references to actions or measures taken

No Comments

No Known Activations