INDEX

Explanations

phrases or terms related to evidence in various contexts

oai_token-act-pair · gpt-3.5-turbo

mentions of evidence in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 0-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.0.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.0.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 ðŁĻĤ

-0.68

iets

-0.67

paces

-0.67

kai

-0.64

star

-0.63

crow

-0.63

[|

-0.62

eon

-0.62

 Runtime

-0.62

rano

-0.62

POSITIVE LOGITS

 evidence

3.58

evidence

3.15

 Evidence

2.78

Evidence

2.77

 proof

2.05

 evid

1.99

 testimony

1.72

 conclusive

1.54

 indications

1.52

 clues

1.44

Activations Density 0.028%

phrases or terms related to evidence in various contexts

mentions of evidence in various contexts

No Comments

No Known Activations