INDEX

Explanations

words related to tools, specifically hammers

oai_token-act-pair · gpt-3.5-turbo

references to hammers and hammering actions

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

RIP

-0.83

VICE

-0.77

ITIES

-0.76

 Charity

-0.72

rees

-0.71

yon

-0.69

Sav

-0.69

ceed

-0.69

ramer

-0.68

Jes

-0.68

POSITIVE LOGITS

 hammer

1.05

sonian

0.97

 hammered

0.94

 hamm

0.94

bats

0.93

heads

0.84

hammer

0.83

 wrench

0.82

matical

0.79

 blow

0.77

Activations Density 0.018%

words related to tools, specifically hammers

references to hammers and hammering actions

No Comments

No Known Activations