INDEX

Explanations

fractional numbers or ratios in text

oai_token-act-pair · gpt-3.5-turbo

phrases indicating statistical occurrences or prevalence

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 4-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.4.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.4.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

-0.74

 disple

-0.65

 nest

-0.61

 arrang

-0.60

âĿ

-0.60

 charism

-0.60

 innoc

-0.59

ens

-0.59

":[

-0.59

 withd

-0.58

POSITIVE LOGITS

organic

1.12

patient

1.06

ordinate

1.04

clusions

1.04

versely

1.03

clus

1.01

animate

1.01

efficiency

1.01

 lieu

1.00

roads

1.00

Activations Density 0.156%

fractional numbers or ratios in text

phrases indicating statistical occurrences or prevalence

No Comments

No Known Activations