INDEX

Explanations

words related to consequences or outcomes

oai_token-act-pair · gpt-3.5-turbo

phrases expressing negation or denial

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.8.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.8.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

Caf

-0.65

 backdrop

-0.62

 Trojan

-0.61

 intersection

-0.61

 Marina

-0.60

 Caucasus

-0.59

 Nicarag

-0.58

geo

-0.58

 partners

-0.57

 beaut

-0.56

POSITIVE LOGITS

iversary

0.89

ember

0.87

heimer

0.77

ï¸ı

0.76

tarian

0.75

0.73

âĤ¬

0.72

Ã§

0.72

scl

0.71

âĻ

0.70

Activations Density 0.110%

words related to consequences or outcomes

phrases expressing negation or denial

No Comments

No Known Activations