INDEX

Explanations

phrases related to persuasive language and statements

oai_token-act-pair · gpt-3.5-turbo

themes related to existential questions and identity

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

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Daytona

-0.72

 Sammy

-0.70

 Speedway

-0.68

 Seah

-0.68

 Rudolph

-0.67

 Hilton

-0.67

 Morty

-0.67

 Kobe

-0.66

 Romo

-0.65

 Canaver

-0.65

POSITIVE LOGITS

='

0.90

?,

0.89

=""

0.77

',

0.77

!'

0.76

=/

0.73

?'

0.72

 doesnt

0.72

speak

0.72

?",

0.72

Activations Density 0.895%

phrases related to persuasive language and statements

themes related to existential questions and identity

No Comments

No Known Activations