INDEX

Explanations

health-related content related to medical conditions and injuries

oai_token-act-pair · gpt-3.5-turbo

instances of numerical or statistical references

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

 inactive

-0.64

 hooked

-0.63

 sometime

-0.60

 append

-0.60

 unsuspecting

-0.59

 totaled

-0.59

Cyp

-0.57

 Fired

-0.56

Ung

-0.56

 Carnegie

-0.56

POSITIVE LOGITS

Speaking

0.83

Recent

0.81

Earlier

0.79

ÂłÂłÂł

0.77

Labour

0.77

Mr

0.76

Ms

0.75

Actor

0.74

Fra

0.73

haw

0.72

Activations Density 0.042%

health-related content related to medical conditions and injuries

instances of numerical or statistical references

No Comments

No Known Activations