INDEX

Explanations

descriptions provided by eyewitnesses

oai_token-act-pair · gpt-3.5-turbo

mentions of eyewitness accounts or testimonies related to events

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

peror

-0.77

enta

-0.70

 legalize

-0.69

ocratic

-0.68

iberal

-0.67

OTUS

-0.66

EEE

-0.65

ãĤª

-0.65

usterity

-0.65

 sail

-0.65

POSITIVE LOGITS

ewitness

1.28

 witnesses

0.96

 eyewitness

0.92

 testimonies

0.87

 corrobor

0.86

Witness

0.86

 testified

0.84

 witness

0.84

 anecd

0.83

 witnessed

0.82

Activations Density 0.334%

descriptions provided by eyewitnesses

mentions of eyewitness accounts or testimonies related to events

No Comments

No Known Activations