INDEX

Explanations

names of specific software or technology platforms

oai_token-act-pair · gpt-3.5-turbo

proper nouns and specific named entities

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 7-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.7.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.7.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

 Jimmy

-0.67

 Stephens

-0.64

aughed

-0.62

¬

-0.61

 Archie

-0.60

ãĤ¸

-0.60

rimp

-0.60

 Herm

-0.59

 Grassley

-0.58

Jimmy

-0.58

POSITIVE LOGITS

 itself

0.80

 experien

0.80

prototype

0.76

's

0.70

 behaves

0.69

 exists

0.66

 Syndrome

0.65

 Goes

0.64

 represents

0.64

 produ

0.64

Activations Density 0.339%

names of specific software or technology platforms

proper nouns and specific named entities

No Comments

No Known Activations