INDEX

Explanations

keywords related to actions or behaviors of people

oai_token-act-pair · gpt-3.5-turbo

references to people and their actions or states

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

 predecessor

-0.96

 srfAttach

-0.95

ãĥ¯

-0.76

 successor

-0.73

 counterpart

-0.70

éŃĶ

-0.70

è¡

-0.69

inth

-0.67

 ausp

-0.66

master

-0.66

POSITIVE LOGITS

 clam

1.09

 complaining

0.88

 dying

0.88

 grav

0.88

 distrust

0.86

 noticing

0.85

 hating

0.85

 starving

0.84

 afraid

0.83

 flock

0.83

Activations Density 0.273%

keywords related to actions or behaviors of people

references to people and their actions or states

No Comments

No Known Activations