INDEX

Explanations

expressions related to emotions and reactions, especially those showing some level of discomfort or disbelief

oai_token-act-pair · gpt-3.5-turbo

intense emotional expressions and reactions

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

arthed

-0.92

sites

-0.73

ricanes

-0.70

ctr

-0.70

senal

-0.69

Sources

-0.69

é¾į

-0.68

 GOODMAN

-0.67

æ©Ł

-0.67

£ı

-0.65

POSITIVE LOGITS

enance

1.15

 grin

1.00

 demeanor

0.97

 smile

0.93

 expression

0.91

 frown

0.86

 smiles

0.85

 innocence

0.84

 Expression

0.84

 laughter

0.83

Activations Density 0.236%

expressions related to emotions and reactions, especially those showing some level of discomfort or disbelief

intense emotional expressions and reactions

No Comments

No Known Activations