INDEX

Explanations

names or word fragments containing "hs."

oai_token-act-pair · gpt-3.5-turbo

mentions of individuals or entities involved in various contexts, possibly relating to significant events or discussions

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 1-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.1.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.1.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

tainment

-0.71

 Trojan

-0.65

 snail

-0.62

gradation

-0.62

 playground

-0.61

 consolid

-0.60

 kinderg

-0.60

 Samoa

-0.59

 fracturing

-0.59

eve

-0.59

POSITIVE LOGITS

hs

1.44

ubs

1.06

omed

0.98

afety

0.96

mith

0.95

ull

0.95

omes

0.91

ttp

0.91

istani

0.89

peed

0.89

Activations Density 0.009%

names or word fragments containing "hs."

mentions of individuals or entities involved in various contexts, possibly relating to significant events or discussions

No Comments

No Known Activations