INDEX

Explanations

words related to performance evaluation and comparison

oai_token-act-pair · gpt-3.5-turbo

terms related to benchmarking and performance evaluation

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

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 5-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.5.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.5.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

yne

-0.78

agan

-0.77

olia

-0.74

oles

-0.69

frey

-0.69

por

-0.69

 Leilan

-0.69

uman

-0.68

uthor

-0.68

ole

-0.68

POSITIVE LOGITS

 benchmarks

1.04

 benchmark

1.02

marks

0.79

 metrics

0.75

bench

0.73

 indices

0.71

 scores

0.71

mark

0.70

 specifications

0.69

 standards

0.69

Activations Density 0.025%

words related to performance evaluation and comparison

terms related to benchmarking and performance evaluation

No Comments

No Known Activations