INDEX

Explanations

evaluative terms expressing personal opinions

oai_token-act-pair · gpt-3.5-turbo

expressions of personal assessment or evaluation

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

perty

-0.64

rench

-0.64

arching

-0.63

idity

-0.61

nance

-0.61

boss

-0.58

ieties

-0.57

orously

-0.56

uador

-0.55

artney

-0.55

POSITIVE LOGITS

 anyway

0.81

 anyways

0.75

,,

0.70

0.69

REDACTED

0.68

 whatsoever

0.67

âĶĢâĶĢâĶĢâĶĢ

0.58

 largeDownload

0.57

 speaking

0.57

elist

0.56

Activations Density 0.102%

evaluative terms expressing personal opinions

expressions of personal assessment or evaluation

No Comments

No Known Activations