INDEX

Explanations

dates and time-related terms

oai_token-act-pair · gpt-3.5-turbo

specific references to time periods or events

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 2-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.2.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.2.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 respectively

-0.61

PRES

-0.59

 fame

-0.57

 vicinity

-0.57

Recomm

-0.54

lease

-0.53

cum

-0.52

 thereof

-0.52

burning

-0.52

Ingredients

-0.52

POSITIVE LOGITS

 aback

0.73

 meanwhile

0.71

endum

0.69

 nutshell

0.69

zsche

0.67

emonium

0.66

hester

0.64

ashtra

0.63

 Explan

0.63

 caveats

0.62

Activations Density 3.215%

dates and time-related terms

specific references to time periods or events

No Comments

No Known Activations