INDEX

Explanations

words indicative of reactions, emotions, or experiences

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

 aeros

-0.68

 disappearing

-0.67

lip

-0.67

 seamlessly

-0.66

 institutional

-0.65

 nicely

-0.64

 dilig

-0.64

 trump

-0.63

awaru

-0.63

 toget

-0.62

POSITIVE LOGITS

numbered

0.76

gins

0.76

ãĤ¦ãĤ¹

0.73

onom

0.72

onomic

0.72

verse

0.71

plex

0.70

ointed

0.69

closed

0.69

lex

0.69

Activations Density 0.047%

words indicative of reactions, emotions, or experiences

No Comments

No Known Activations