INDEX

Explanations

bike-related terms

oai_token-act-pair · gpt-3.5-turbo

references to modes of transportation and related activities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

FORE

-0.74

 ende

-0.74

etsk

-0.73

nces

-0.73

iqueness

-0.72

rences

-0.71

ternity

-0.70

grave

-0.69

ertodd

-0.69

 situ

-0.68

POSITIVE LOGITS

 ladder

0.97

 broom

0.93

wagon

0.92

 wagon

0.88

 treadmill

0.87

roller

0.86

 stairs

0.85

 roller

0.84

 raft

0.81

rollers

0.80

Activations Density 0.122%

bike-related terms

references to modes of transportation and related activities

No Comments

No Known Activations