INDEX

Explanations

phrases related to political inquiries and public messaging

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

soever

-0.72

ageddon

-0.68

agically

-0.68

imately

-0.66

 Says

-0.65

etimes

-0.62

 Hence

-0.59

abus

-0.59

 Perhaps

-0.59

 dictated

-0.59

POSITIVE LOGITS

Mos

0.70

ajor

0.69

ÄŁ

0.67

mos

0.66

perty

0.66

Ro

0.65

 hairc

0.62

umbn

0.61

pport

0.61

OTOS

0.60

Activations Density 7.969%

phrases related to political inquiries and public messaging

No Comments

No Known Activations