INDEX

Explanations

statements providing information or explaining concepts

oai_token-act-pair · gpt-3.5-turbo

sentences that assert the existence or importance of ideas and concepts

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

wi

-0.75

deen

-0.74

ebus

-0.74

culosis

-0.72

someone

-0.70

cue

-0.69

was

-0.69

keeper

-0.69

 Debate

-0.68

erva

-0.66

POSITIVE LOGITS

 grouped

1.12

 legion

1.00

 reflections

0.98

 collectively

0.96

 ones

0.94

 staples

0.92

 clustered

0.92

 interchangeable

0.92

 numbered

0.92

 adaptations

0.92

Activations Density 0.349%

statements providing information or explaining concepts

sentences that assert the existence or importance of ideas and concepts

No Comments

No Known Activations