INDEX

Explanations

mentions of public speaking events or interviews

oai_token-act-pair · gpt-3.5-turbo

the word "at" representing various contexts of location or time

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.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.11.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

irez

-0.72

usercontent

-0.69

uay

-0.67

activate

-0.65

ister

-0.63

REDACTED

-0.59

alties

-0.59

uploads

-0.58

escent

-0.58

 survives

-0.58

POSITIVE LOGITS

 length

1.02

 conferences

0.96

 least

0.91

CES

0.89

SX

0.88

 rallies

0.87

 halftime

0.85

TED

0.84

CP

0.84

 Cannes

0.81

Activations Density 0.107%

mentions of public speaking events or interviews

the word "at" representing various contexts of location or time

No Comments

No Known Activations