INDEX

Explanations

names of actors and phrases related to law or legal actions

oai_token-act-pair · gpt-3.5-turbo

names of celebrities and notable figures

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

)."

-0.74

 muster

-0.69

upon

-0.68

 culminating

-0.67

 latter

-0.63

sic

-0.62

'."

-0.60

espie

-0.59

 keen

-0.59

 sadly

-0.58

POSITIVE LOGITS

 Doesn

1.04

Â¶

0.98

 Yourself

0.97

 Versus

0.89

Isn

0.87

 Gets

0.85

 Makes

0.85

 Approach

0.84

 Wouldn

0.84

 Costs

0.84

Activations Density 0.360%

names of actors and phrases related to law or legal actions

names of celebrities and notable figures

No Comments

No Known Activations