INDEX

Explanations

references to power and control in societal contexts

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

mean

-0.78

sov

-0.77

isode

-0.76

borough

-0.71

Pen

-0.69

æĺ

-0.67

Fel

-0.67

bent

-0.66

cloth

-0.66

furt

-0.66

POSITIVE LOGITS

 independently

0.73

 metab

0.67

rooms

0.67

 multiplier

0.66

 pedals

0.65

 conscience

0.64

 interstate

0.64

 unlimited

0.63

 STATS

0.61

 bulldo

0.61

Activations Density 0.505%

references to power and control in societal contexts

No Comments

No Known Activations