INDEX

Explanations

mentions of the term 'ba'

oai_token-act-pair · gpt-3.5-turbo

references to the term "ba" in various contexts

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

andem

-0.77

MENT

-0.72

ALE

-0.71

Uncommon

-0.70

lessly

-0.68

IOC

-0.68

ivities

-0.68

serious

-0.67

lich

-0.65

 Editors

-0.64

POSITIVE LOGITS

Ã±

0.99

ÅĤ

0.92

uble

0.90

 fide

0.83

uthor

0.83

atar

0.82

ffe

0.80

ques

0.79

ilee

0.79

ÅŁ

0.79

Activations Density 0.012%

mentions of the term 'ba'

references to the term "ba" in various contexts

No Comments

No Known Activations