INDEX

Explanations

the word "that" in various contexts

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs49152-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_49152

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

torch.float32

Hook Point

blocks.8.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

ogether

-0.72

ãĤ©

-0.68

bledon

-0.62

erenn

-0.61

 redes

-0.60

onder

-0.59

oufl

-0.59

istani

-0.58

Guard

-0.56

rily

-0.56

POSITIVE LOGITS

[+

0.67

 doesnt

0.60

 they

0.60

 Allaah

0.59

ihad

0.58

 there

0.56

we

0.55

 although

0.55

 RELEASE

0.55

ndra

0.54

Activations Density 0.161%

the word "that" in various contexts

No Comments

No Known Activations