INDEX

Explanations

occurrences of specific symbols or punctuation marks

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res_post_32k-oai

Configuration

jbloom/GPT2-Small-OAI-v5-32k-resid-post-SAEs/v5_32k_layer_10.pt

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

32,768

Data Type

torch.float32

Hook Name

blocks.10.hook_resid_post

Hook Layer

Architecture

standard

Context Size

Dataset

Skylion007/openwebtext

Activation Function

topk

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 afore

-0.67

 bonuses

-0.64

 irony

-0.63

 Ober

-0.63

 Leia

-0.63

 alternatives

-0.62

 glue

-0.62

 transitional

-0.62

 accelerated

-0.61

 emulation

-0.61

POSITIVE LOGITS

minute

1.38

year

1.38

month

1.37

hour

1.37

mile

1.34

hr

1.31

week

1.30

unit

1.27

acre

1.26

lb

1.25

Activations Density 0.170%

occurrences of specific symbols or punctuation marks

No Comments

No Known Activations