INDEX

Explanations

negations and phrases indicating the absence of something

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_post_32k-oai

Configuration

jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_8.pt

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

32,768

Data Type

torch.float32

Hook Name

blocks.8.hook_resid_post

Hook Layer

Architecture

standard

Context Size

Dataset

Skylion007/openwebtext

Activation Function

topk

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

emo

-0.74

ULT

-0.64

tests

-0.64

xt

-0.63

testing

-0.63

ITNESS

-0.63

door

-0.60

value

-0.60

bum

-0.60

paper

-0.60

POSITIVE LOGITS

offic

0.77

 Lowry

0.71

 ausp

0.67

ITION

0.66

ãĥ¤

0.65

liga

0.64

anke

0.64

adel

0.63

surprisingly

0.63

Led

0.63

Activations Density 0.333%

negations and phrases indicating the absence of something

No Comments

No Known Activations