INDEX

Explanations

CEO names and related terms

oai_token-act-pair · gpt-3.5-turbo

references to CEOs and their positions or actions

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

 compr

-0.71

 needles

-0.71

adoes

-0.69

wikipedia

-0.69

theless

-0.69

merga

-0.67

tro

-0.66

-0.65

Prev

-0.65

icultural

-0.65

POSITIVE LOGITS

IAL

0.82

 Larry

0.76

HAEL

0.76

doms

0.73

ynthesis

0.71

 Markus

0.70

 Alvin

0.70

 Warren

0.69

 Mansion

0.68

owered

0.68

Activations Density 0.028%

CEO names and related terms

references to CEOs and their positions or actions

No Comments

No Known Activations