INDEX

Explanations

news-related terms or headlines

oai_token-act-pair · gpt-3.5-turbo

statements characterized by strong adjectives or descriptors

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

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

assies

-0.71

rooms

-0.70

verty

-0.69

 fame

-0.69

 corridors

-0.69

Frames

-0.67

æ©

-0.67

 formations

-0.67

 constituencies

-0.67

SPONSORED

-0.65

POSITIVE LOGITS

 spokesperson

0.84

ccording

0.83

 researcher

0.77

 example

0.75

 spokesman

0.71

 staffer

0.70

 spokeswoman

0.69

 refres

0.66

 exception

0.66

 therapist

0.65

Activations Density 0.467%

news-related terms or headlines

statements characterized by strong adjectives or descriptors

No Comments

No Known Activations