INDEX

Explanations

words related to sudden, often violent, outbursts or conflicts

oai_token-act-pair · gpt-3.5-turbo

instances of conflict or confrontational situations

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

ittee

-0.76

ajo

-0.75

ndra

-0.72

 ease

-0.67

leans

-0.65

thin

-0.65

ileen

-0.65

avement

-0.65

abled

-0.65

eline

-0.63

POSITIVE LOGITS

burst

0.76

 bursting

0.74

 anew

0.73

 shells

0.71

 Spiegel

0.70

 spontaneously

0.68

ULAR

0.66

 Kiev

0.66

 uncontroll

0.65

ONSORED

0.65

Activations Density 0.040%

words related to sudden, often violent, outbursts or conflicts

instances of conflict or confrontational situations

No Comments

No Known Activations