INDEX

Explanations

words related to the concept of "iconic" or "iconicity."

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

Sph

-0.75

 unfocusedRange

-0.67

 shutter

-0.66

ãĤ°

-0.66

 blazing

-0.65

Nex

-0.65

 Shogun

-0.64

DERR

-0.63

enance

-0.62

 Blazing

-0.60

POSITIVE LOGITS

arb

0.96

ourt

0.93

apped

0.88

opter

0.87

cup

0.86

itta

0.84

ott

0.82

asia

0.80

hene

0.80

ap

0.80

Activations Density 0.013%

words related to the concept of "iconic" or "iconicity."

No Comments

No Known Activations