INDEX

Explanations

phrases related to personal opinions and reflections

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

Detects the lexical token for thinking/cognition (the verb and its appearances in multiword phrases and compounds).

oai_token-act-pair · gpt-5-mini Triggered by @forsubscription1

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

eatures

-0.74

idated

-0.73

benef

-0.66

andro

-0.64

adra

-0.63

enges

-0.63

{"

-0.63

rance

-0.62

Benef

-0.62

estial

-0.60

POSITIVE LOGITS

 toget

0.66

 provoking

0.63

ashi

0.63

 aloud

0.62

xual

0.62

 spont

0.61

pad

0.60

 impart

0.58

aults

0.58

ovie

0.57

Activations Density 0.260%

phrases related to personal opinions and reflections

Detects the lexical token for thinking/cognition (the verb and its appearances in multiword phrases and compounds).

No Comments

No Known Activations