INDEX

Explanations

phrases related to names, titles, and medical terms

oai_token-act-pair · gpt-3.5-turbo

specific names and terms related to individuals or entities in a context

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.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.2.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

BILITIES

-0.71

 chains

-0.66

Hopefully

-0.65

rieg

-0.63

ngth

-0.62

HAEL

-0.61

Attempts

-0.60

 sidx

-0.60

 SHARES

-0.60

 DEFENSE

-0.59

POSITIVE LOGITS

ysis

0.79

QL

0.75

ilver

0.72

oleon

0.70

zsche

0.66

 anymore

0.63

bledon

0.63

iqueness

0.62

heastern

0.62

EMBER

0.62

Activations Density 0.213%

phrases related to names, titles, and medical terms

specific names and terms related to individuals or entities in a context

No Comments

No Known Activations