INDEX

Explanations

words related to physical entities or actions related to trenches

oai_token-act-pair · gpt-3.5-turbo

references to trenches

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

Ring

-0.79

visual

-0.77

ghai

-0.76

ivia

-0.74

isphere

-0.74

estate

-0.72

bah

-0.71

cius

-0.70

iesel

-0.69

isc

-0.68

POSITIVE LOGITS

anty

0.87

coat

0.83

 Warfare

0.78

rower

0.76

 warfare

0.75

 Hare

0.71

 frontier

0.70

ãĥ¤

0.69

auld

0.67

 Colossus

0.65

Activations Density 0.050%

words related to physical entities or actions related to trenches

references to trenches

No Comments

No Known Activations