INDEX

Explanations

positions or titles followed by phrases related to organizations or institutions

oai_token-act-pair · gpt-3.5-turbo

references to various official titles and roles

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

Joined

-0.72

theless

-0.67

scale

-0.67

dayName

-0.65

 doses

-0.64

phrine

-0.64

 plaus

-0.64

 suscept

-0.63

 sche

-0.63

 ancest

-0.63

POSITIVE LOGITS

orney

0.92

querque

0.88

 Audit

0.74

medi

0.74

wered

0.71

 Warwick

0.71

 Programming

0.71

 Communication

0.70

forestation

0.70

 Strategic

0.69

Activations Density 0.068%

positions or titles followed by phrases related to organizations or institutions

references to various official titles and roles

No Comments

No Known Activations