INDEX

Explanations

punctuation and specific named entities or places

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 3-att_32k-oai

Configuration

jbloom/GPT2-Small-OAI-v5-32k-attn-out-SAEs/v5_32k_layer_3

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

32,768

Data Type

torch.float32

Hook Name

blocks.3.hook_attn_out

Hook Layer

Architecture

standard

Context Size

Dataset

Skylion007/openwebtext

Activation Function

topk

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 tremend

-0.70

 unanswered

-0.69

 conclud

-0.68

 embold

-0.63

 rewarded

-0.61

 gauge

-0.61

redict

-0.60

ãĤ¤ãĥĪ

-0.60

arya

-0.59

KT

-0.59

POSITIVE LOGITS

nant

0.74

gypt

0.71

wives

0.69

clair

0.65

 Crescent

0.65

onga

0.65

ca

0.64

resso

0.63

andro

0.63

Mig

0.61

Activations Density 0.830%

punctuation and specific named entities or places

No Comments

No Known Activations