INDEX

Explanations

phrases indicating the author's personal opinion

oai_token-act-pair · gpt-3.5-turbo

expressions of personal opinions and views

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ancial

-0.74

 Brief

-0.65

 Beaut

-0.62

bard

-0.62

uddin

-0.59

 Start

-0.59

artney

-0.59

 deposited

-0.58

Submit

-0.57

 Starts

-0.57

POSITIVE LOGITS

Īè

0.79

·

0.77

©¶æ¥µ

0.72

naissance

0.71

§

0.70

ħĭ

0.69

©¶æ

0.69

uyomi

0.68

etheless

0.68

ĸ

0.68

Activations Density 0.073%

phrases indicating the author's personal opinion

expressions of personal opinions and views

No Comments

No Known Activations