INDEX

Explanations

verbal phrases indicating information or updates being shared

oai_token-act-pair · gpt-3.5-turbo

verbs indicating belief, announcement, or information disclosure

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

isode

-0.71

 vulner

-0.63

apeake

-0.61

quin

-0.60

Exit

-0.59

stad

-0.58

vette

-0.58

lua

-0.58

ston

-0.57

ohan

-0.57

POSITIVE LOGITS

 that

0.93

that

0.83

by

0.76

 advisable

0.67

 unanimously

0.64

 amongst

0.62

BY

0.61

uality

0.59

 THAT

0.58

 incidentally

0.58

Activations Density 0.057%

verbal phrases indicating information or updates being shared

verbs indicating belief, announcement, or information disclosure

No Comments

No Known Activations