INDEX

Explanations

instances of the word "inv" or "involved" in the text

oai_token-act-pair · gpt-3.5-turbo

words related to invasion or incursion

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 10-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.10.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.10.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

Ħ¢

-0.66

©¶æ

-0.66

 bare

-0.66

 blending

-0.61

 Tale

-0.61

 untreated

-0.59

 disson

-0.59

 horizontally

-0.59

 Whole

-0.59

 hump

-0.59

POSITIVE LOGITS

igor

1.47

incible

1.46

asion

1.46

ocation

1.45

itational

1.44

olve

1.43

olved

1.43

ocations

1.42

itations

1.41

itation

1.37

Activations Density 0.019%

instances of the word "inv" or "involved" in the text

words related to invasion or incursion

No Comments

No Known Activations